Division Does Not Imply Predictability: Demographics Continue to Reveal Little About Voting and Partisanship

Abstract

What are the political consequences of ongoing social sorting? We evaluate the degree of social sorting and mass polarization using the predictability of partisanship and voting decisions as quantities of interest. Contrary to expectations, demographic sorting has not produced a very predictable electorate. Models trained on nothing more than demographic labels from public opinion surveys (1952–2020) predict only 63.9% of two-party vote choices and 63.4% of partisan IDs correctly out-of-sample—whether they be based on logistic regressions or tree-based machine learning models. Moreover, demographics’ predictive power over vote choice or partisan affiliation shows a surprising stability over time. We argue that while select demographics’ marginal effects may appear to be evidence of social sorting, the joint predictability of political behavior using only demographic characteristics has been, and still is, modest at best.